Identifying Simplified Models For Load Shifting Control

Buildings do not only make up a major part of the energy consumption of industrial countries, modern buildings also have a lot of heating, air-conditioning and cooling (HVAC) equipment, which uses electric energy to supply thermal energy for indoor comfort. Most of this equipment is over-engineered to cope with peaks, aging and losses, and therefore is operated in duty cycling most of the time. This setup can be exploited for the sake of optimization. Building simulation tools (e.g. ENERGY+, TRNSYS) can take the complete physical behavior of a complex building into account, but are computationally intensive. They also rely on detailed geometrical models of the building, which causes costs for modeling and thus makes it hard to implement simulation as the enabler for large scale model-based load shifting. Since buildings and their HVAC systems have quite slow time dynamics, a feasible approach is doing worst case estimations for the thermal behavior off-line and using this data for the on-line control of the HVAC system.